Blab AI
AI Algorithm Detects Trafficked Marine Wildlife in Airport X-Rays
Researchers have released the first artificial‑intelligence system that can identify trafficked dead marine species from 3D X‑ray scans. The study, published last week in Frontiers in Ocean Sustainability, shows the algorithm can detect shark fins, seahorses and sea cucumbers with high accuracy.
The work was led by marine biologist Vanessa Pirotta of Macquarie University. According to the paper, the team collected 68 samples of the three species from the Australian Museum Collections, a beached bull shark, and a grocery store in Sydney. They then created about 6,000 simulated luggage bags, 3,500 of which contained hidden marine parts and 2,400 that did not. The bags were scanned with 3D X‑ray machines supplied by Rapiscan, a U.S. security‑equipment manufacturer that also funded the research.
The algorithm achieved 95–96 % detection rates for shark fins and seahorses and 86 % for sea cucumbers. False‑alarm rates were 9 % for seahorses and 1–2 % for the other two species. Pirotta noted that the system is a “complementary tool” and “not a silver bullet.” She added that the algorithm’s performance could be limited by the variation among sea cucumbers and by the need for 3D X‑ray machines that can produce real‑time images.
Marine wildlife trafficking is a growing but under‑reported problem. Interpol’s 2025 seizure data show that more marine specimens were confiscated than reptiles, birds and primates combined, with 91,000 pieces seized last year. The trade is driven by demand for ornamental fish, luxury foods and traditional medicines, and most animals are moved in airplane luggage or airmail, where they often go undetected.
The algorithm’s focus on small airborne luggage means it does not address sea‑freight shipments, a gap highlighted by researchers. Anagnostou, a fellow at Oxford’s illegal‑wildlife‑trade program, emphasized that detection is only the first step in a longer chain that requires forensics and prosecution.
In Argentina, authorities seized a large shipment of dead fish, octopuses and crabs at an airport near Buenos Aires on April 26, 2026. The incident, reported by the Associated Press, was the third time in a year that the same airport had intercepted illegal marine wildlife. The rescue centre had to install ten new emergency tanks to care for survivors.
The study’s authors argue that the algorithm could help airport security and mail‑handling facilities flag suspicious items for further inspection. They plan to release the algorithm’s “recipe” so that other regions can adapt it to additional species.
The work follows a 2022 paper by the same team that used 3D X‑ray imaging to detect terrestrial wildlife trafficking. The new marine‑species algorithm builds on that methodology but requires a larger dataset of marine samples and more sophisticated image‑analysis techniques.
The researchers acknowledge that the system’s effectiveness depends on the availability of 3D X‑ray scanners at airports and on the ability of inspectors to act on flagged bags. They also note that the algorithm can only work with dead, mostly dried samples, and that distinguishing species within the same genus remains a challenge.
The study was welcomed by enforcement agencies, but officials cautioned that technology alone cannot stop trafficking. A spokesperson for the United Nations Office on Drugs and Crime said, “Detection is the first link in a longer chain, not the whole answer.” The algorithm is intended to assist human inspectors and biosecurity dogs, not replace them.
Marine wildlife trafficking also poses ecological and public‑health risks. Illegal trade can spread infectious diseases, introduce invasive species and fuel other forms of organized crime. The International Union for Conservation of Nature notes that fish are often overlooked in conservation discussions, which focus more on terrestrial species.
The algorithm’s development involved an interdisciplinary team that considered how traffickers conceal animal parts. The researchers created realistic smuggling scenarios by hiding marine parts in toys, clothing and tin foil, mirroring common concealment tactics.
While the system shows promise, its practical deployment will require widespread access to suitable X‑ray equipment, training for security staff, and integration with existing inspection workflows. The authors suggest that future work could expand the species database and improve performance on live or fresh specimens.
In summary, the new AI algorithm represents a first step toward automated detection of trafficked marine wildlife at airports and mail facilities. Its high accuracy for shark fins, seahorses and sea cucumbers demonstrates the potential of 3D X‑ray imaging combined with machine‑learning techniques. However, the technology is not a standalone solution; it must be paired with human expertise, legal frameworks and broader enforcement strategies to curb the growing trade in marine animals.
The work was led by marine biologist Vanessa Pirotta of Macquarie University. According to the paper, the team collected 68 samples of the three species from the Australian Museum Collections, a beached bull shark, and a grocery store in Sydney. They then created about 6,000 simulated luggage bags, 3,500 of which contained hidden marine parts and 2,400 that did not. The bags were scanned with 3D X‑ray machines supplied by Rapiscan, a U.S. security‑equipment manufacturer that also funded the research.
The algorithm achieved 95–96 % detection rates for shark fins and seahorses and 86 % for sea cucumbers. False‑alarm rates were 9 % for seahorses and 1–2 % for the other two species. Pirotta noted that the system is a “complementary tool” and “not a silver bullet.” She added that the algorithm’s performance could be limited by the variation among sea cucumbers and by the need for 3D X‑ray machines that can produce real‑time images.
Marine wildlife trafficking is a growing but under‑reported problem. Interpol’s 2025 seizure data show that more marine specimens were confiscated than reptiles, birds and primates combined, with 91,000 pieces seized last year. The trade is driven by demand for ornamental fish, luxury foods and traditional medicines, and most animals are moved in airplane luggage or airmail, where they often go undetected.
The algorithm’s focus on small airborne luggage means it does not address sea‑freight shipments, a gap highlighted by researchers. Anagnostou, a fellow at Oxford’s illegal‑wildlife‑trade program, emphasized that detection is only the first step in a longer chain that requires forensics and prosecution.
In Argentina, authorities seized a large shipment of dead fish, octopuses and crabs at an airport near Buenos Aires on April 26, 2026. The incident, reported by the Associated Press, was the third time in a year that the same airport had intercepted illegal marine wildlife. The rescue centre had to install ten new emergency tanks to care for survivors.
The study’s authors argue that the algorithm could help airport security and mail‑handling facilities flag suspicious items for further inspection. They plan to release the algorithm’s “recipe” so that other regions can adapt it to additional species.
The work follows a 2022 paper by the same team that used 3D X‑ray imaging to detect terrestrial wildlife trafficking. The new marine‑species algorithm builds on that methodology but requires a larger dataset of marine samples and more sophisticated image‑analysis techniques.
The researchers acknowledge that the system’s effectiveness depends on the availability of 3D X‑ray scanners at airports and on the ability of inspectors to act on flagged bags. They also note that the algorithm can only work with dead, mostly dried samples, and that distinguishing species within the same genus remains a challenge.
The study was welcomed by enforcement agencies, but officials cautioned that technology alone cannot stop trafficking. A spokesperson for the United Nations Office on Drugs and Crime said, “Detection is the first link in a longer chain, not the whole answer.” The algorithm is intended to assist human inspectors and biosecurity dogs, not replace them.
Marine wildlife trafficking also poses ecological and public‑health risks. Illegal trade can spread infectious diseases, introduce invasive species and fuel other forms of organized crime. The International Union for Conservation of Nature notes that fish are often overlooked in conservation discussions, which focus more on terrestrial species.
The algorithm’s development involved an interdisciplinary team that considered how traffickers conceal animal parts. The researchers created realistic smuggling scenarios by hiding marine parts in toys, clothing and tin foil, mirroring common concealment tactics.
While the system shows promise, its practical deployment will require widespread access to suitable X‑ray equipment, training for security staff, and integration with existing inspection workflows. The authors suggest that future work could expand the species database and improve performance on live or fresh specimens.
In summary, the new AI algorithm represents a first step toward automated detection of trafficked marine wildlife at airports and mail facilities. Its high accuracy for shark fins, seahorses and sea cucumbers demonstrates the potential of 3D X‑ray imaging combined with machine‑learning techniques. However, the technology is not a standalone solution; it must be paired with human expertise, legal frameworks and broader enforcement strategies to curb the growing trade in marine animals.
